230 Dr. Andrews on the Heat developed 



certain number of degrees, was suddenly cooled by Immersion in it. Denoting 

 by t the loss of heat sustained by the thermometer, and by If the temperature 

 gained by the liquid, I obtained in different trials the following numbers : 



12 3 



t = 59°.00, t = 69°.00, t = 72°.00. 



t'= o°.9o, tf = r.oo, if= 1M5. 



Hence, we deduce for the value of the thermometer in grammes of water, 



12 3 Mean. 



0.47, 0.45, 0.49, 0.47. 



6. From the last two results we may therefore conclude, that the brass vessel 

 and thermometer, taken together, are equivalent to 1.09 gr. water. 



7. A very important source of error in this and other similar investigations, 

 where the variation of temperature of a liquid requires to be observed with the 

 utmost precision, arises from the cooling influence of the surrounding air during 

 the time occupied by the observation, which, in the experiments I am about to 

 describe, amounted to nearly 1'. Where the increase of temperature does not 

 exceed 2° or 3° Fah., the common method of cooling the liquid before the 

 experiment begins, as many degrees below the temperature of the air as it will 

 afterwards rise above it, may be employed with success ; but for greater incre- 

 ments of heat, this process is liable to a serious error, which it is necessary to 

 avoid. In fact, on mixing the liquids together, the thermometer attains, in 

 a very few seconds, almost its ultimate point of elevation, and it occupies at 

 least four-fifths of the entire time in rising through the last half degree. As, 

 therefore, the mixture continues much longer in the upper than in the lower 

 half of its range of temperature, the method just described will necessarily yield 

 results sensibly below the truth.* In practice, this error may be effectually 

 obviated, by reducing the initial temperature of the liquid so far below the 

 temperature of the air, that its final maximum may never reach higher than 

 2° F. above the same point. 



• A similar observation has been made by M. Regnault in his recent and valuable memoir on 

 the " Specific Heats of Simple and Compound Bodies" (Ann. de Chin. t. 63, p. 23) ; but the error 

 thus induced he corrects by means of an interpolating formula. 



